GEOPHYSICAL RESEARCH LETTERS, VOL. 30, NO. 10, 1506, doi:10.1029/2002GL016274, 2003
The N:Si:P molar ratio in the Strait of Gibraltar
Evgeny V. Dafner,1 Roberta Boscolo,2 and Harry L. Bryden3
Received 12 September 2002; accepted 13 March 2003; published 20 May 2003.
[1] All existing descriptions of nutrient distributions in the
Strait of Gibraltar suggest that the Atlantic water brings to
the Mediterranean Sea nutrients in the Redfield ratio (N:Si:P
= 16:15:1). Here, the N:Si:P molar ratios (±Standard Error),
obtained in April 1998, are used to show that in the Atlantic
water at the western entrance of the Strait this ratio is lower
(13.8(±0.5):12.1(±1.0):1) than the classical Redfield ratio; it
is close to the Redfield ratio in the middle of the Strait
(15.6(±0.6):10.7(±0.9):1), and increases dramatically to
23.6(±3.4):29.1(±4.5):1 at the eastern entrance of the
Strait. In the Mediterranean water, the N:Si:P ratio has a
quite similar trend with 31.5(±6.0):26.5(±3.6):1 in the east,
20.4(±0.2):31.5(±11.1):1 in the middle and 18.1(±0.6):
17.6(±0.7):1 in the west of the Strait. The physical and
biological processes that account for the observed spatial
variability of the N:Si:P ratio along the Strait are identified.
We estimated that in the Atlantic water entering the
Mediterranean Sea, about 84% of the variability in N:Si:P
molar ratio is due to biological and 16% to physical
INDEX TERMS: 4845 Oceanography: Biological
processes.
and Chemical: Nutrients and nutrient cycling; 4283 Oceanography:
General: Water masses; 4805 Oceanography: Biological and
Chemical: Biogeochemical cycles (1615); KEYWORDS: nutrient,
Redfield molar ratio, the Strait of Gibraltar. Citation: Dafner, E.
V., R. Boscolo, and H. L. Bryden, The N:Si:P molar ratio in the
Strait of Gibraltar, Geophys. Res. Lett., 30(10), 1506, doi:10.1029/
2002GL016274, 2003.
1. Introduction
[2] The molar ratio between nitrate, silicate and phosphate in marine phytoplankton and in deep oceanic waters is
generally constant (N:Si:P = 16:15:1) and is known as the
Redfield ratio [Redfield et al., 1963]. In this ratio phytoplankton consume nutrients from seawater, and bacteria
mineralize organic nutrients to inorganic. One of the unresolved paradoxes of the Mediterranean Sea is the high
N:Si:P ratio in deep waters: 24:32:1 in the Eastern and
22:19.5:1 in the Western Basins [Béthoux et al., 2002].
Nitrate, silicate and phosphate concentrations in the Mediterranean Sea are controlled by the exchange through the
Straits of Gibraltar and Bosphorus, by the atmospheric
deposition and by river discharge.
[3] To explain these high ratios we need to understand
the balance of nutrients in the Mediterranean Sea. The
1
Department of Oceanography, School of Ocean and Earth Science and
Technology, University of Hawaii, Honolulu, Hawaii, USA.
2
CSIC - Instituto de Investigaciones Mariñas, Vigo, Spain.
3
Southampton Oceanography Center, Empress Dock, Southampton,
UK.
Copyright 2003 by the American Geophysical Union.
0094-8276/03/2002GL016274
13
Atlantic water is one component of this balance that is still
poorly understood [Béthoux et al., 1998]. Despite the fact
that the Atlantic water entering the Mediterranean Sea is
depleted in nutrients [e. g., Minas et al., 1991; Turley,
1999], all existing descriptions of nutrients in the Strait
suggest that these nutrients are related in the Redfield ratio
[Béthoux et al., 2002]. Most of these observations were
made on the western entrance of the Strait, and they were
applied to the Atlantic water in the Mediterranean Sea
[Béthoux et al., 2002].
[4] High Redfield ratios (30 to 50) found in the Alboran
Sea - Algerian Current have been interpreted as a signature of earlier nutrient uptake that has taken place within a
high N:P environment [Raimbault and Coste, 1990].
Minas and Minas [1995] have emphasized that the water
circulation near Europa Point on Gibraltar and in the
Alboran Sea determines recycling of nutrients in the
Mediterranean Sea, which maintains the high N:P tendency over the whole basin. Here we show that N:Si:P
ratio in the Atlantic water in the eastern entrance of the
Strait is higher than 16:15:1, and that this ratio increases
dramatically from the western to the eastern entrance of
the Strait. In the Mediterranean water the N:Si:P ratio has
a quite similar behavior.
2. Material and Methods
[5] Samples were collected in the Strait of Gibraltar area
between 12 and 18 April 1998 on board of the RRV
Discovery, cruise 232 (Figure 1). A CTD rosette system
(Seabird) equipped with 10 L Niskin bottles was used.
Stations were chosen to cover a variety of waters from
western (in the Spartel sill area) and eastern (in the Western
Alboran Sea) entrances of the Strait, and in the middle of
the Strait (Tarifa Narrows). The bottles were fired at the
depths of maximum or minimum distribution of temperature, salinity, oxygen and fluorescence, within and between
different water bodies and at the interface layer between the
Atlantic inflow and Mediterranean outflow. Samples were
also taken at few meters from the bottom. A detailed
description of the depth profiles and bottled sampling are
shown in Dafner et al. [2001].
[6] The Winkler whole bottle titration method with
amperometric endpoint detection for dissolved oxygen
determination was used [Culberson and Huang, 1987].
Water samples for nitrate plus nitrite (hereafter referred as
nitrate), silicate, and phosphate analysis were collected after
CFC, dissolved oxygen and total organic carbon samples.
Chemical constituents were measured on board immediately
after sampling. All nutrients were analyzed with a Chemlab
Auto Analyser II coupled to a Digital-Analysis Microstream
data capture and reduction system [Hydes, 1984]. Specification of the chemical analyses, including precision, accuracy, and detection limits are provided in Table 1. In this
- 1
13 - 2
DAFNER ET AL.: N:SI:P RATIO IN THE STRAIT OF GIBRALTAR
Figure 1. Location of sampling stations in the Strait of
Gibraltar during cruise 232 on board of the RRV Discovery
(12 – 18 April, 1998). To the east of the Strait lies the
Alboran Sea, the western-most basin of the Mediterranean
Sea. To the west of the Strait lies the Gulf of Cádiz, an
embayment of the Northeast Atlantic Ocean.
work, data on phytoplankton composition and biomass were
taken from [Gómez et al., 2000].
3. Results and Discussion
[7] The study area and the hydrographic conditions found
in the Strait of Gibraltar in April 1998 have been already
described in greater detail [Dafner et al., 2001]. The Redfield molar ratios based on individual nutrient concentrations (Figure 2, second line) suggest that the N:Si:P ratios in
the Atlantic water at the western entrance of the Strait are
lower than the classical Redfield ratio. In the middle of the
Strait the N:P ratio is close to the Redfield ratio and Si:P
ratio is lower than the Redfield ratio; both of these ratios
increase considerably at the eastern entrance of the Strait.
The Redfield ratios calculated with the use of the average
nutrient concentrations (Figure 2, third line) show that the
N:Si:P ratios are close to the Redfield values in the west and
middle of the Strait, but they are significantly higher in the
east of the Strait. The discrepancies between two calculations can be attributed to the variability in nutrient concentrations affected by physical (upwelling) and biological
(phytoplankton consumption) processes along the Eurasian
and African coasts of the Strait. Standard errors of the ratios
calculated from the individual nutrient measurements show
higher variability in values to the east of the Strait than to
the west. These variations in the Atlantic water to the east of
the Strait are due to the upwelling of intermediate waters
and nutrient utilization by phytoplankton.
Table 1. The Overall Performance of Dissolved Oxygen (mmol
kg 1) and Nutrients Analyses (mM) Based on Laboratory Tests
Constituent
Detection
limit
Precision
Oxygen
Silicate
Phosphate
Nitrate
2.0
0.05
0.02
0.05
0.6
0.05
0.02
0.05
Accuracy at typical
concentration
0.6 mmol kg 1 at 7.8 mmol kg
1.5 mM at 100 mM
0.05 mM at 2.0 mM
0.05 mM at 30.0 mM
1
Figure 2. The sketch of the two-layers model of water
mass exchange through the Strait of Gibraltar with average
concentrations of nitrate, silicate and phosphate (in mM,
±SD, first line), N:Si:P molar ratio (±Standard Error) based
on the calculation of individual nutrient concentrations
(second line) and average nutrient concentrations for each
transect (third line), and number of measurements, and
salinity values used for mixing analysis calculated for each
transect as an average for the entire water mass (fourth line).
Values of Standard Error in line two are calculated from
molar ratio values. Entrainment transports between the
Atlantic and Mediterranean waters are shown as WTR (1Sv
= 106 m3 s 1), and nitrate, silicate and phosphate exchanges
between these layers as NTR, SiTR and PTR, respectively.
The Atlantic and the Mediterranean waters are separated by
the interface layer, which is identified by salinity values of
37.0, 37.32 and 37.5 on the western, middle and eastern
entrance of the Strait, respectively. The depth of this layer
increases from about 130– 160 m on the southwest to about
15– 20 m on the northeast.
[8] In the surface (Atlantic) water, the decrease in the
N:Si ratio from the middle of the Strait to the eastern
entrance (1.46 to 0.81, respectively), indicates a higher
consumption of silicate relative to nitrate by phytoplankton.
Additionally, oxygen over-saturation in the Atlantic water
(up to 5%) and phosphate concentration below the detection
limit (<0.02 mM, data not shown) suggest that despite the
short length of the Strait (60 km) and a highly dynamic
environment (surface currents can reach 3 – 4 knots), these
conditions are favorable for photosynthesis by phytoplankton. The chlorophyll values and microphytoplankton biomass increase from the southwest towards the northeast of
the Strait, in agreement with the ascent of the interface layer.
The microphytoplankton assemblage in central and eastern
parts of the Strait is dominated by diatoms (>90% of total
biomass), and is very different from the phytoplankton
assemblage found to the west of the Strait, which has a
dominance of dinoflagellates and relatively high concentration of microheterotrophs [Gómez et al., 2000]. Figure 2
shows that in the Gulf of Cádiz, concentrations of silicate
are lower (first line) than at Tarifa Narrows and in the
Western Alboran Sea. It is well known that the lack of
13 - 3
DAFNER ET AL.: N:SI:P RATIO IN THE STRAIT OF GIBRALTAR
silicate may, to some extend, determine species succession
from a diatom to a flagellate community [Parsons et al.,
1984], and finally can be responsible for the differences in
the N:Si ratio.
[9] In the deep (Mediterranean) water, the N:Si:P molar
ratio exhibits similar behavior to that of the Atlantic water:
the ratio decreases from east to west. Close to the eastern
entrance of the Strait at a depth of about 300 m, an oxygen
minimum and nitrate and phosphate concentration maxima
(data not shown) indicate an active mineralization of
organic material by bacteria. The Alboran Sea oxygen
minimum and nutrient maximum are related to increased
productivity in the surface layer associated with upwelling
in the northern sector of the Alboran anticyclonic gyre
[Packard et al., 1988; Minas et al., 1991]. The regeneration of organic material at this depth in the Western
Alboran Sea increases the N:P ratio to values significantly
higher (31.5(±6.0):1, Figure 2) than those found in the
Eastern Mediterranean (24:1, [Béthoux et al., 2002]). After
mixing in the Strait with the Atlantic water, the Mediterranean water leaves the Strait with a ratio closer to the
Redfield ratio (Figure 2). The phytoplankton and bacterial
activities significantly increase the N:Si:P ratio in the
surface and deep waters at the eastern entrance of the
Strait. Data from both the northwestern [Thingstad et al.,
1998] and the eastern Mediterranean [Zohary and Robarts,
1998] indicate that the growth rate not only of phytoplankton, but also of heterotrophic bacteria, is phosphoruslimited during stratification.
[10] One physical phenomenon that affects the biology in
the Strait is the strong shear between Atlantic and Mediterranean waters in the Camarinal Sill area where the
Atlantic water accelerates and entrains the Mediterranean
water [Wesson and Gregg, 1994]. In order to estimate the
shear from the mass conservation equation applied to the
two layers system in the Strait, we used the water transport
values obtained during the Gibraltar [Bryden et al., 1994;
Bray et al., 1995] and CANIGO [Tsimplis and Bryden,
2000] experiments. There are several different water transport estimates in the Strait of Gibraltar in the literature,
however we believe that the estimates from the aforementioned experiments are the most accurate and precise
because they are based on a large yearlong data set.
[11] From the difference between inflow and outflow
between two layers, we estimated that the entrainment of
the Atlantic water at the western entrance of the Strait is
0.15 Sv (0.84 Sv – 0.69 Sv), while the value for the
Mediterranean water at the eastern entrance is 0.13 Sv
(0.69 Sv – 0.82 Sv; Figure 2). The first value is slightly
higher than the 0.1 Sv that has been presented by Wesson
and Gregg [1994] to the east of the Camarinal Sill. For the
Atlantic water in the west and Mediterranean water in the
east, we compared the rate of entrainment with the rate of
water transport for each water mass, i.e. 0.15 Sv with
0.84 Sv in the west, and 0.13 Sv with 0.82 Sv on the
east, respectively. These results show that 18% of the
Atlantic water with ‘low’ N:Si:P ratio re-circulates back to
the Atlantic with Mediterranean water and 16% of the
Mediterranean water with ‘high’ N:Si:P ratio flows back
to the Mediterranean with Atlantic water. The uncertainty of
these estimates is associated with the accuracy of the water
transport estimates, which according to the CANIGO obser-
Table 2. N, Si and P Values (mM) of the Mediterranean Water
(MW) in the Gulf of Cádiz and the Atlantic Surface Water (ASW)
in Western Alboran Sea Determined From Two-Point Mixture
Calculations
Water masses
Sal
N
N
Si
Si
P
P
MW
ASW
37.627
36.432
7.7
4.8
1.3
2.0
6.8
3.0
1.9
0.3
0.36
0.30
0.14
0.17
N, Si and P are the differences between the average observed values
(first line in Figure 2) and those calculated from mixing analysis (in mM).
vations is about 0.06 Sv. This accuracy implies that variations of the entrainment between inflow and outflow can be
as high as 40% for the re-circulation of the Atlantic water
and 46% for the re-circulation of the Mediterranean water.
[12] We find higher nutrient concentrations in the Atlantic water to the east of the Camarinal Sill than to the west
illustrating the input of nutrient constituents due to the
mixing between two layers (Figure 2). It is well documented
that when mixing events occur at or near the sill region of
the Strait, the Atlantic water is enriched with nutrients and
advected towards the Mediterranean [e.g., Minas et al.,
1991; Echevarrı́a et al., 2002]. Minas and Minas [1995]
noted that Mediterranean water induces a hydrological
recycling of high N:P water at the sill, back to the Mediterranean. The very high N:P waters (20 to 50) are situated
just under the halocline in the Alboran Sea from where they
are partly destined to go back to the western Mediterranean
basin.
[13] Semidiurnal tidal fluctuations in the Strait can
reverse the flows [La Violette and Lacombe, 1988], and
the periodic occurrence of nonlinear internal waves can
support mixing that increases the N:Si:P ratio in the Atlantic
and decreases it in the Mediterranean waters. Nonlinear
internal wave packets are generated at the main bathymetric
sill (the Camarinal sill) located in the western approaches to
the Strait and have been found 200 km inside the Alboran
Sea [Pistek and La Violette, 1999].
[14] Additionally, close to Europa Point on Gibraltar
there is an upwelling that brings deep waters with ‘high’
N:Si:P ratios to the surface (Figure 2). The upwelling is an
almost permanent feature in this area. Two mechanisms
have been suggested for the upwelling dynamic in this
region: wind stress and the southward drifting of the
Atlantic water [Tarek et al., 2000]. Wind-driven upwelling
dominates in coastal zones and on the shelf, while upwelling associated with southward drifting of the Atlantic water
prevails further offshore. Wind-driven upwelling influences
a larger area adjacent to the Strait and lifts water that has a
higher nutrient concentration [Tarek et al., 2000].
[15] Mixing analysis has been performed on the mean
nutrient values given in Figure 2 (first line). By considering
the values at the top left of Figure 2 as the values of ‘‘pure’’
Atlantic water and the values at bottom right for ‘‘pure’’
Mediterranean water and using salinity as a conservative
component, we calculated the 2-point mixture values of N,
Si and P for the other two water masses at the west and east
ends of the Strait, i.e. Mediterranean and Atlantic waters
respectively. We first applied the linear mixing equation to
the salinity values (see Table 2) in order to derive the
percentage of Mediterranean and Atlantic water mixed by
physical processes. Using these mixing coefficients, the
13 - 4
DAFNER ET AL.: N:SI:P RATIO IN THE STRAIT OF GIBRALTAR
values of N, Si and P resulting from physical processes were
calculated from the ‘‘pure water’’ values of N, Si, and P
(Table 2). The differences between the observed and calculated values quantify to what extent the biological processes
affect the water within the Strait. The differences in Table 2
indicate that the nutrient content of the Mediterranean water
in Gulf of Cádiz is modified by remineralization, while
consumption is affecting the nutrient concentrations of the
Atlantic water in the western Alboran Sea. In the Atlantic
water the differences between observed values and those
from mixing analysis are negative while in the Mediterranean water, concentrations of observed nutrients are lower
than those predicted from the mixing model.
[16] Using the vertical and horizontal nutrient fluxes in
the Strait (Figure 2), we estimated that in the Atlantic water
entering the Mediterranean Sea, physical and biological
processes contribute about 16% and 84%, respectively, to
the N:Si:P increase. More observations on the mass transport due to upwelling in the area close to Europa Point
would help to clarify the contribution of biological processes. The Atlantic water entering the Mediterranean Sea is
already depleted of nutrients, especially phosphate, with
N:Si:P ratios similar to the deep waters of the Eastern and
Western Basins. These observations are in disagreement
with the conclusion by Krom et al. [1991] that states that the
degree of phosphorus limitation in the Mediterranean Sea
increases from west to east across the entire basin. The
result presented here implies that the role of the Atlantic
water in the nutrient balance of the Mediterranean Sea
should be reconsidered.
CSIC) are acknowledged as they helped to improve the current version of
the manuscript. This paper is contribution number 277 from the Center for
Marine Sciences, UNCW (for ED).
References
[17] Spatial distribution of nutrients in the Strait of
Gibraltar during April 1998 shows that the N:Si:P molar
ratios in the Atlantic water entering the Strait (western
entrance) are lower than the classical Redfield ratios and
increase dramatically along the Strait going from the western to the eastern entrance of the Mediterranean. A similar
increase is found for the outflowing Mediterranean waters.
The ratios in the surface and deep waters at the eastern
entrance of the Strait are similar to those found in the
Eastern Mediterranean Basin, thus suggesting that the role
of the Atlantic water in the nutrient balance of the Mediterranean Sea should be reconsidered. The biological processes responsible for the observed spatial variability are
nutrient consumption by phytoplankton and nutrient regeneration by bacteria in the surface and deep waters at the
eastern entrance of the Strait. However physical processes
like vertical shear at the interface of the Atlantic and
Mediterranean waters, tidal fluctuations and upwelling are
identified as the main phenomena affecting the biology in
the Strait. Finally, our results indicate that the Strait of
Gibraltar area is phosphorus-limited for the growth of
phytoplankton and heterotrophic bacteria.
Béthoux, J. P., P. Morin, and D. P. Ruiz-Pino, Temporal trends in nutrient
ratios: chemical evidence of Mediterranean ecosystem changes driven by
human activity, Deep-Sea Res. II, 49(11), 2007 – 2016, 2002.
Béthoux, J. P., P. Morin, C. Chaumery, O. Connan, B. Gentili, and D. RuizPino, Nutrients in the Mediterranean Sea, mass balance and statistical
analysis of concentrations with respect to environmental change, Mar.
Chem., 63, 155 – 169, 1998.
Bray, N. A., J. Ochoa, and T. N. Kinder, The role of the interface in
exchange through the Strait of Gibraltar, J. Geophys. Res., 100(C6),
10,755 – 10,776, 1995.
Bryden, H. L., J. Candela, and T. H. Kinder, Exchange through the Strait of
Gibraltar, Prog. Oceanogr., 33, 201 – 248, 1994.
Culberson, C. H., and S. Huang, Automated amperometric oxygen titration,
Deep-Sea Res., 34, 875 – 880, 1987.
Dafner, E. V., R. Sempéré, and H. L. Bryden, Total organic carbon distribution and budget through the Strait of Gibraltar in April 1998, Mar.
Chem., 73, 223 – 252, 2001.
Echevarrı́a, F., J. G. Lafuente, M. Bruno, G. Gorsky, M. Goutx, N. González, C. M. Garcı́a, F. Gómez, J. M. Vargas, M. Picheral, L. Striby,
M. Varela, J. J. Alonso, A. Reul, A. Cózar, L. Prieto, T. Sarhan, F. Plaza,
and F. Jiménez-Gómez, Physical-biological coupling in the Strait of Gibraltar, Deep-Sea Res. II, 49(19), 4115 – 4130, 2002.
Gómez, F., N. González, F. Echevarrı́a, and C. M. Garcı́a, Distribution and
fluxes of dissolved nutrients in the Strait of Gibraltar and its relationship
to microplankton biomass, Est. Coast. and Shelf Sci., 51, 439 – 449, 2000.
Hydes, D. J., A manual of methods for the continuous flow determination
of nutrients in seawater, IOSDL report 177, 40 pp., 1984.
Krom, M. D., N. Kress, and L. I. Gordon, Phosphorus limitation of primary
productivity in the eastern Mediterranean Sea, Limn. Oceanogr., 36(3),
424 – 432, 1991.
La Violette, P. E., and H. Lacombe, Tidal-induced pulses in the flow
through the Strait of Gibraltar, Oceanolog. Acta, SP9, 13 – 27, 1988.
Minas, H. J., and M. Minas, Consideration on the Mediterranean Redfield
ratio anomalies with special attention to oxygen vs nutrient changes. The
role of water circulation, Wat. Pol. Res., 107 – 123, 1995.
Minas, H. J., B. Coste, P. Le Corre, M. Minas, and P. Raimbault, Biological
and geochemical signatures associated with the water circulation through
the Strait of Gibraltar and in the Western Alboran Sea, J. Geophys. Res.,
96(C5), 8755 – 8771, 1991.
Packard, T. T., et al., Formation of the Alboran oxygen minimum zone,
Deep-Sea Res., 35(7), 1111 – 1118, 1988.
Parsons, T. R., M. Takahashi and B. Hargrave, Biological oceanographic
processes, pp. 330, Pergamon Press, Oxford, 1984.
Pistek, P., and P. E. La Violette, Observations of the suppression of tidegenerated nonlinear internal wave packets in the Strait of Gibraltar,
J. Mar. Syst., 20, 113 – 128, 1999.
Raimbault, P., and B. Coste, Very high values of the nitrate:phosphate ratio
(>30) in the subsurface layers of the western Mediterranean Sea, Rapp.
Comm. Int. Mer. Medit., 32(1), C-18, 1990.
Redfield, A. C., B. H. Ketchum and F. A. Richards, The influence of
organisms on the composition of seawater, In The Sea, edited by M. N.
Hill, vol. 2, pp. 26 – 77, Interscience, New York, 1963.
Tarek, S., J. Garcı́a Lafuente, M. Vargas, J. M. Vargas, and F. Plaza, Upwelling mechanisms in the northwestern Alboran Sea, J. Mar. Sys., 23(4),
317 – 331, 2000.
Thingstad, T. F., U. L. Zweifel, and F. Rossoulzadegan, P-limitation of both
phytoplankton and heterotrophic bacteria in NW Mediterranean summer
surface waters, Limnol. Oceanogr., 43, 88 – 94, 1998.
Tsimplis, M., and H. L. Bryden, Estimation of transport through the Strait
of Gibraltar, Deep-Sea Res., 47(12), 2219 – 2242, 2000.
Turley, C. M., The changing Mediterranean Sea - a sensitive ecosystem?
Progr, Oceanogr., 44, 387 – 400, 1999.
Wesson, J. C., and M. C. Gregg, Mixing at Camarinal Sill in the Strait of
Gibraltar, J. Geophys. Res., 99(C5), 9847 – 9878, 1994.
Zohary, T., and R. D. Robarts, Experimental study of microbial p-limitation
in the eastern Mediterranean, Limn. Oceanogr., 43, 387 – 395, 1998.
[18] Acknowledgments. This research was funded by the European
Commission, MAST III Programme (Contract MAS3 - CT96 - 0060).
Shiptime aboard RRV Discovery was provided by the Natural Environment
Research Council under the core strategic research Programme ‘‘Largescale, Long-term Ocean Circulation’’ at Southampton Oceanography
Centre. Financial support for E.V. Dafner came from Ministere Affaires
Etrangères Francais and Conseil Général des Bouches du Rhône, France.
The comments of an anonymous reviewer and Dr. Carmen G. Castro (IIM-
E. V. Dafner, Department of Oceanography, School of Ocean and Earth
Science and Technology, University of Hawaii, Honolulu, HI 96822,
U.S.A. ([email protected])
R. Boscolo, CSIC - Instituto de Investigaciones Mariñas, Eduardo
Cabello 6, 36208, Vigo, Spain.
H. L. Bryden, Southampton Oceanography Center, Empress Dock,
Southampton, SO14 3ZH, U.K.
4. Summary